PERSPECTIVES IN DIABETES Sugar, Uric Acid, and the Etiology of Diabetes and Obesity Richard J. Johnson,1,2 Takahiko Nakagawa,1,3 L. Gabriela Sanchez-Lozada,4 Mohamed Shafiu,5 Shikha Sundaram,6 Myphuong Le,1 Takuji Ishimoto,1 Yuri Y. Sautin,7 and Miguel A. Lanaspa1

The intake of added sugars, such as from table sugar (sucrose) For example, a high intake of fructose induces leptin re- and high-fructose corn syrup has increased dramatically in the sistance in rats (7). Fructose also encourages food intake last hundred years and correlates closely with the rise in obesity, due to stimulation of dopamine in the mesolimbic system metabolic syndrome, and diabetes. Fructose is a major compo- and effects on the hypothalamus (8,9). Food intake is also nent of added sugars and is distinct from other sugars in its ability stimulated by hepatic ATP depletion (10), which occurs in to cause intracellular ATP depletion, nucleotide turnover, and the animals and humans administered fructose (11). Fructose generation of uric acid. In this article, we revisit the hypothesis may also affect metabolic rate. A recent study in humans that it is this unique aspect of fructose metabolism that accounts documented a reduction in resting energy expenditure for why fructose intake increases the risk for metabolic syn- drome. Recent studies show that fructose-induced uric acid in overweight and obese subjects fed fructose but not generation causes mitochondrial oxidative stress that stimulates glucose (12). fat accumulation independent of excessive caloric intake. These studies challenge the long-standing dogma that “a calorie is just FRUCTOSE-INDUCED METABOLIC SYNDROME DOES NOT a calorie” and suggest that the metabolic effects of food may REQUIRE INCREASED ENERGY INTAKE matter as much as its energy content. The discovery that fructose- mediated generation of uric acid may have a causal role in diabetes The ability for fructose (and sucrose, which contains and obesity provides new insights into pathogenesis and therapies fructose) to stimulate food intake and to lower metabolism for this important disease. Diabetes 62:3307–3315, 2013 provides a mechanism for how a high fructose intake may encourage weight gain and visceral fat accumulation. However, fructose or sucrose also alters fat stores and metabolism independent of excessive energy intake. Al- though weight gain is largely controlled by overall energy FRUCTOSE-INDUCED WEIGHT GAIN AND METABOLIC intake, other features of metabolic syndrome can occur SYNDROME independent of weight gain. For example, rats fed fructose develop fatty liver, hypertriglyceridemia, and insulin re- Experimental studies from the 1950s showed the peculiar sistance when compared with rats fed isocaloric glucose ability of fructose to induce insulin resistance in laboratory or starch-enriched diets (4,5). Indeed, hypertriglyceridemia, rats. Today, fructose intake has been shown to induce all fatty liver, and type 2 diabetes can be induced in metabolic features of metabolic syndrome in rats, as well as oxi- syndrome–prone rats with caloric restriction provided the dative stress, endothelial dysfunction, fatty liver, micro- fi diet is high (40%) in sucrose (which contains fructose) (5). albuminuria and kidney disease (rev. in 1). Similar ndings A recent epidemiological analysis in humans also found an can be shown when animals are fed sucrose or high-fructose association of diabetes prevalence with sugar availability corn syrup (HFCS), both which contain fructose (2,3). In that was independent of total energy intake (13). contrast, administration of glucose or starch results in fewer features of metabolic syndrome when provided equivalent intake (4,5). A ROLE FOR URIC ACID IN FRUCTOSE-INDUCED FAT Fructose may increase the risk for obesity by altering ACCUMULATION satiety, resulting in increased food intake. The intake of The observation that fructose-fed rats develop fatty liver fructose is not effective in stimulating insulin and leptin and metabolic syndrome without requiring increased en- secretion in humans, and hence may not induce a satiety ergy intake suggests that the metabolism of fructose may response (6). Other mechanisms may also be operative. be different from that of other carbohydrates. Fructose is distinct from glucose only in its initial metabolism. The first enzyme to metabolize fructose is fructokinase (also From the 1Division of Kidney Diseases and Hypertension, University of Colo- known as ketohexokinase [KHK]). The metabolism of rado Denver, Aurora, Colorado; the 2Division of Nephrology, Eastern Colo- rado Health Care System, Department of Veteran Affairs, Denver, Colorado; fructose to fructose-1-phosphate by KHK occurs primarily the 3TMK Project, Medical Innovation Center, Kyoto University, Kyoto, Ja- in the liver, is rapid and without any negative feedback, pan; the 4Laboratory of Renal Physiopathology and Department of Nephrol- and results in a fall in intracellular phosphate and ATP ogy, Instituto Nacional de Cardiologia I.Ch., Mexico City, Mexico; 5Renal 6 – Associates, San Antonio and Del Rio, Texas; the Division of Pediatric Gas- levels (14 16). This has been shown to occur in the liver troenterology, Children’s Hospital, Aurora, Colorado; and the 7Division of in humans with relatively small doses of oral fructose (60 Nephrology, Hypertension and Renal Transplantation, University of Florida, gfructosealoneor39gfructosewith39gglucose)(11). Gainesville, Florida. Corresponding author: Richard J. Johnson, [email protected]. The decrease in intracellularphosphatestimulatesAMP Received 1 February 2013 and accepted 17 June 2013. deaminase (AMPD), which catalyzes the degradation of DOI: 10.2337/db12-1814 AMP to inosine monophosphate and eventually uric acid Ó 2013 by the American Diabetes Association. Readers may use this article as (15) (Fig. 1). The increase in intracellular uric acid is long as the work is properly cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons.org/licenses/by followed by an acute rise in uric acid in the circulation -nc-nd/3.0/ for details. likely due to its release from the liver (14). Fructose also diabetes.diabetesjournals.org DIABETES, VOL. 62, OCTOBER 2013 3307 FRUCTOSE, URIC ACID, AND DIABETES

As such, this is an interesting condition in which marked nucleotide turnover and ATP depletion occur but without the ability to be further metabolized by this primary en- zymatic pathway to glucose, glycogen, or triglycerides (20). Nevertheless, the feeding of fructose to HepG2 cells lacking aldolase B resulted in a rapid accumulation of triglycerides, consistent with our findings that uric acid itself can induce triglyceride accumulation (19). These experiments explain why fatty liver and hyperuricemia are common complications of this disease (21) and also why fatty liver and diabetes are complications in subjects with glycogen storage disease I, in which hepatic intracellular ATP depletion and hyperuricemia also occur (22–25). Fi- nally, it provides an explanation for why fructose is lipo- genic (based on acetate labeling studies) despite little of the fructose molecule being incorporated into the tri- glyceride molecule itself (19). We next addressed how the degradation of nucleotides might lead to fat accumulation. Specifically, our group and others have shown that AMPD counters the effects of AMP-activated protein kinase (AMPK) (26,27). Whereas activation of AMPK in hepatocytes induces oxidation of fatty acids and ATP generation, AMPD has opposite effects. Overexpression of AMPD in HepG2 cells blocks fatty acid oxidation and increases fat accumulation, whereas silencing AMPD blocks fructose-induced fat ac- cumulation. The mechanism is mediated in part by the generation of uric acid, which inhibits AMPK (27). In addition to inhibiting AMPK, uric acid may stimulate FIG. 1. Fructose-induced nucleotide turnover. Fructose is rapidly hepatic lipogenesis (28). The mechanism appears to be me- phosphorylated in the hepatocyte by KHK to fructose-1-phosphate (F-1-P), which uses ATP as a phosphate donor. Intracellular phosphate (PO4) diated by uric acid–dependent intracellular and mitochon- levels decrease, stimulating the activity of AMP deaminase 2 (AMPD2). drial oxidative stress (28). Although uric acid is a potent AMPD2 converts AMP to inosine monophosphate (IMP). IMP is me- antioxidant in the extracellularenvironment,whenuricacid tabolized to inosine by 59 nucleotidase (59NT), which is further de- fi graded to xanthine and hypoxanthine by xanthine oxidase (XO), enters cells via speci corganicaniontransporters, it induces ultimately generating uric acid. an oxidative burst that has been shown in vascular smooth muscle cells, endothelial cells, adipocytes, islet cells, renal tubular cells, and hepatocytes (29–31). Uric acid–induced stimulates uric acid synthesis from amino acid pre- oxidative stress appears to be mediated by the stimulation cursors, such as glycine (17). of NADPH oxidase, which translocates to mitochondria Recent studies suggest that this “side event” in fructose (28,29,32). Uric acid can also generate triuretcarbonyl and metabolism may be critical for how fructose induces aminocarbonyl radicals as well as alkylating species upon metabolic syndrome. First, there are actually two KHK reaction with peroxynitrite and can also directly inactivate isoforms, and they differ in their ability to activate this nitric oxide (NO) to 6-aminouracil (33,34). pathway. KHK-C phosphorylates fructose rapidly, con- The induction of oxidative stress in the mitochondria suming ATP with the generation of uric acid. In contrast, causes a reduction in aconitase-2 activity in the Krebs KHK-A phosphorylates fructose slowly and consumes cycle, resulting in citrate accumulation that is transported minimal ATP (18). When both KHK-C and KHK-A are de- into the cytoplasm where it activates ATP citrate lyase, leted, mice are fully protected from fructose-induced acetyl CoA carboxylase, and fatty acid synthase, leading to metabolic syndrome and fatty liver (18); however, when fat synthesis (19). Uric acid also causes a reduction in KHK-A is selectively deleted, there is increased fructose enoyl CoA hydratase-1, a rate-limiting enzyme in b-fatty available for metabolism by KHK-C, and the metabolic acid oxidation (35). The consequence is fat accumulation syndrome and fatty liver are worsened compared with wild- in the hepatocyte (19,35). type mice despite the same intake of total calories and fruc- Recently, we identified another mechanism by which tose (18). These studies suggest that differences in nucleotide uric acid may increase the risk for hepatic fat accumula- turnover might influence the metabolic response. tion and metabolic syndrome. Fructose (or sucrose) in- To examine the purine nucleotide pathway in the met- gestion is known to increase hepatic KHK levels (5). The abolic response, we silenced aldolase B in a hepatocyte increased expression of KHK is driven in part by the pro- line (HepG2 cells) (19). Aldolase B is the second enzyme duction of uric acid from fructose (35) A rise in in- in fructose metabolism, and the genetic loss of aldolase tracellular uric acid activates the nuclear transcription B is the cause of hereditary fructose intolerance. When factor, carbohydrate responsive element–binding protein aldolase B is inhibited, fructose is phosphorylated by ATP (35). When KHK expression is increased in HepG2 cells by but cannot be further metabolized, nevertheless fructose can uric acid exposure, the triglyceride accumulation in re- be metabolized by other routes such as hexokinase. In this sponse to fructose is enhanced (35). regard, subjects with hereditary fructose intolerance, are This is relevant to subjects with nonalcoholic fatty known to have hyperactive KHK and show enhanced ATP liver disease (NAFLD). Subjects with NAFLD ingest depletion and uric acid generation in response to fructose. more fructose-containing soft drinks than age, sex, and

3308 DIABETES, VOL. 62, OCTOBER 2013 diabetes.diabetesjournals.org R.J. JOHNSON AND ASSOCIATES

explain how fructose induces insulin resistance. Mito- chondrial oxidative stress has a role in driving insulin re- sistance (42). In turn, the development of fatty liver is also linked with insulin resistance (43). Effects in the white adipose tissue. Uric acid may also induce insulin resistance via effects on adipocytes. Uric acid is taken up in adipocytes by an organic anion trans- porter where it induces oxidative stress via activation of NADPH oxidase, generating oxidized lipids and inflammatory mediators such as monocyte chemoattractant protein-1 (MCP-1) (29,44). Adiponectin synthesis is also inhibited (44). In the hyperuricemic Pound mouse, the inhibition of uric acid synthesis by allopurinol attenuates the local inflammatory response in the visceral fat, reduces the expression of inflammatory cytokines, and enhances circu- lating levels of adiponectin in association with an improve- FIG. 2. Classic and alternative lipogenic pathways of fructose. In the ment in insulin resistance (44). Likewise, the reduction of classical pathway, triglycerides (TG) are a direct product of fructose uric acid by either allopurinol or benzbromarone in the metabolism by the action of multiple enzymes including aldolase B (Aldo B) and fatty acid synthase (FAS). An alternative mechanism was fructose-fed rat results in less insulin resistance and recently shown (30). Uric acid produced from the nucleotide turnover decreases the leptin overexpression that occurs in the that occurs during the phosphorylation of fructose to fructose-1-phosphate visceral fat (4,45). (F-1-P) results in the generation of mitochondrial oxidative stress (mtROS), which causes a decrease in the activity of aconitase (ACO2) Vascular effects. Fructose may also induce insulin re- in the Krebs cycle. As a consequence, the ACO2 substrate, citrate, accu- sistance via effects on the vasculature. One of the major mulates and is released to the cytosol where it acts as substrate for TG effects of insulin is to stimulate the release of NO from synthesis through the activation of ATP citrate lyase (ACL) and fatty acid synthase. AMPD2, AMP deaminase 2; IMP, inosine monophosphate; PO4, endothelial cells, where it causes vasodilation that aids phosphate. delivery of glucose to the skeletal muscle. Mice that cannot generate endothelial NO develop features of metabolic syndrome and insulin resistance (46). In this regard, uric acid inhibits endothelial NO generation, including in re- BMI-matched control subjects and have increased KHK ex- sponse to insulin (32). Uric acid reduces endothelial NO pression in their liver (36). Subjects with NAFLD who have via several mechanisms, including blocking the uptake of the highest fructose intake also show the greatest ATP de- the substrate, L-arginine (47), stimulating the degradation pletion in response to a fructose load, and those subjects with of L-arginine by arginase (48), and scavenging NO by uric the highest uric acid levels show a greater nadir in the ATP acid or by uric acid–generated oxidants (32,34,49). Hyper- depletion (37). These data are consistent with an induction of uricemic rats have impaired endothelial function and hy- KHK in the liver with subsequent increased sensitivity to the pertension that can be reversed by lowering uric acid or effects of fructose via a uric acid–dependent mechanism. treating with L-arginine or antioxidants (50–52). Hyperuri- Rodents have lower serum uric acid than humans due to cemia is also associated with endothelial dysfunction in the presence of uricase in their liver, and hence show humans, and lowering uric acid with allopurinol improves a lesser rise in serum uric acid in response to fructose (38). endothelial dysfunction in asymptomatic hyperuricemia, Nevertheless, lowering uric acid has also been found to congestive heart failure, diabetes, chronic kidney disease, block the development of hepatic steatosis in fructose-fed obstructive sleep apnea, and with smoking (rev. in 53). rats (35). Lowering uric acid also reduces hepatic steatosis Islet cell effects. Chronic administration of fructose or in the desert gerbil (which spontaneously develops fatty sucrose to animals not only causes insulin resistance but liver on a normal diet) (39), in alcoholic fatty liver (in which may also result in type 2 diabetes (5,54). Histologically, the increased intrahepatic uric acid occurs) (40), and in the islets show hyalinosis and macrophage infiltration, similar Pound mouse (a mouse model of metabolic syndrome to what is observed in humans with type 2 diabetes. The manifesting fatty liver, obesity, insulin resistance, and hy- mechanism by which fructose induces these changes is not pertension caused by a leptin receptor mutation) (28). known because the islet does not express GLUT5, which is These studies supported the tight association of hyperuri- the primary fructose transporter. However, we reported an cemia with fatty liver; prospective studies have also reported upregulation of the urate transporter URAT-1 in islet cells that an elevated uric acid independently predicts the de- of sucrose-fed rats in association with increased expres- velopment of NAFLD (41). The ability of hyperuricemia to sion of MCP-1 (5). Incubation of cultured insulin-secreting predict fatty liver is independent of obesity. Hyperuricemia islet cells with uric acid also causes oxidative stress and is even associated with NAFLD in hemodialysis subjects synthesis of MCP-1 (5). Oxidative stress in islets is con- who have a BMI below 20 (19). A summary of how fructose sidered to have a major role in causing the islet dysfunc- and uric acid induce fatty liver is shown in Fig. 2. tion of type 2 diabetes.

FRUCTOSE-INDUCED HYPERURICEMIA, INSULIN RESISTANCE, AND DIABETES EVIDENCE THAT FRUCTOSE MEDIATES FATTY LIVER The observation that inhibition of uric acid synthesis pre- AND INSULIN RESISTANCE IN HUMANS vented metabolic syndrome and hepatic steatosis leads to The major source of fructose in the Western diet is from the question of how uric acid might contribute to insulin soft drinks and fruit drinks, and this accounts for ap- resistance and diabetes. proximately 7% of caloric intake in the adult, and upward Hepatic effects. The observation that uric acid can in- to 15% of total caloric intake in adolescents. Intake of sugar duce mitochondrial oxidative stress and fatty liver may and soft drinks are higher in populations at increased risk diabetes.diabetesjournals.org DIABETES, VOL. 62, OCTOBER 2013 3309 FRUCTOSE, URIC ACID, AND DIABETES for insulin resistance and diabetes, including the African and nucleotide synthesis. Initially the rise in serum uric Americans, Hispanics, Native Americans, and subjects with acid is best shown shortly (30–60 min) after fructose in- lower income. A meta-analysis concluded that the intake of gestion (or ingestion of HFCS or sucrose), but total 24-h sugary soft drinks is an independent predictor for the de- levels are also elevated (60,68). Over time, fasting serum velopment of metabolic syndrome and/or diabetes (55). uric acid levels increase (58). Intake of soft drinks is also Genetic factors enhance the risk for developing diabetes associated with increasing risk for hyperuricemia (69). from soft drinks (56). An elevated serum uric acid is also one of the best in- Clinical studies have documented the metabolic effects dependent predictors of diabetes and commonly precedes of fructose. Studies from the 1960s through the 1980s the development of both insulin resistance and diabetes showed that sucrose, or fructose, can worsen hyper- (Table 1). An elevated uric acid also independently pre- triglyceridemia and insulin resistance, especially if subjects dicts the development of fatty liver (41), obesity (70), hy- were hyperinsulinemic (57,58). More recently Stanhope pertension (rev. in 71), and elevations in C-reactive protein et al. (59) fed 25% of diet as fructose or glucose to over- (72). Furthermore, metabolic syndrome is associated with weight individuals for 10 weeks. Although some features a high frequency of hyperuricemia, and similarly, hyper- of metabolic syndrome were induced with glucose, the uricemia is associated with metabolic syndrome (73,74). fructose-fed subjects showed worse postprandial hyper- Though hyperinsulinemia may contribute to hyperuricemia triglyceridemia, increased hepatic de novo synthesis of fatty by blocking uric acid excretion, it cannot be the primary acids, a decrease in insulin sensitivity (noted by elevations reason for the association because hyperuricemia has in fasting glucose and insulin levels), increased total and been reported to precede the development of hyper- visceral fat (among men), higher 24-h uric acid levels, in- insulinemia and/or diabetes (Table 1). creased systemic inflammatory mediators (MCP-1), and A number of conditions associated with hyperuricemia lower resting energy expenditure (12,59,60). In another are also associated with increased risk for insulin re- study, Maersk et al. (61) randomized overweight adults to sistance or diabetes, including chronic lead intoxication drink 1 L of a sugary soft drink daily for 6 months, with and gestational diabetes mellitus. Many drugs associated control subjects receiving equivalent amount of diet soft with insulin resistance are also associated with hyperuri- drink, milk, or water. At the end of 6 months, the subjects cemia, such as calcineurin inhibitors and thiazide di- receiving the sugary soft drinks displayed more visceral, uretics. Indeed, lowering uric acid improves the insulin skeletal muscle, and liver fat and higher serum trigly- resistance induced by thiazides in rats (75). cerides and cholesterol compared with the group drink- Evidence that lowering uric acid can improve insulin ing milk, with a trend toward significance in the other resistance in humans is limited. One small study showed two groups. Tappy and colleagues (62) have also shown that lowering uric acid with benzbromarone improves in- the ability of fructose to induce insulin resistance, hepatic sulin resistance in subjects with congestive heart failure lipid accumulation, and hypertriglyceridemia. Similarly, our (76). Another study reported that lowering uric acid group administered 200 g fructose to overweight men for improves HbA1c levels in normotensive diabetic subjects 2weeksanddocumentedhigherbloodpressure,higher (77). In contrast, we were not able to show an improve- triglycerides, and lower HDL cholesterol compared with ment of insulin resistance with allopurinol in subjects ad- baseline, with 25% of the subjects developing de novo ministered fructose (63), but the doses of fructose were metabolic syndrome at 2 weeks (63). Another study showed exceptionally high (200 g/day) raising the possibility that that the administration of one 8-oz soft drink per day to the doses of allopurinol we used might not have been able adolescents results in increased body weight at 18 months to block intracellular uric acid. Clearly, more studies are compared with subjects given diet soft drinks (64). indicated before any definitive conclusions can be made Intervention studies have also been performed to eval- with regards to the benefit of lowering uric acid for the uate the effect of reducing sugar intake on metabolic treatment of insulin resistance. syndrome. For example, the Atkins diet and other low carbohydrate diets tend to improve features of the meta- bolic syndrome more than typical low fat diets (65). PROBLEMS WITH THE FRUCTOSE AND URIC ACID A randomized study in school children reported that re- HYPOTHESIS ducing soft drink intake, resulting in a difference of 175 Concerns with animal studies. The fructose-induced mL/day between treatment and control subjects, led to hyperuricemia hypothesis has been challenged. First, ani- a reduction in overweight or obesity by 0.2% in the treated mal studies using fructose typically use pure fructose as group compared with a 7.5% increase in the control sub- opposed to sucrose or HFCS, which is the primary source jects at 12 months (66). A study in California showed that of fructose in humans, and the dose of fructose adminis- the banning of soft drinks in schools resulted in a re- tered to rodents is usually higher (50–60% of the diet) duction in overall soft drink intake with a decrease in compared with humans (where it is typically 10–15% of the obesity in children 6 to 11 years of age (67). Less effect diet). Purified fructose is used, however, so one can sep- was observed in older children, possibly because the arate the effects of fructose from glucose. Indeed, animals overall reduction in soft drink intake in this latter group are more sensitive to the combination of fructose and was less effective (67). Soft drink intake in the U.S. has glucose because both sugars accelerate the absorption of decreased since peaking in 1999, and this is also associ- the other (78). Combinations of free fructose and glucose, ated with a leveling of the rates of obesity. or sucrose, induce features of metabolic syndrome with levels of fructose of 20–30% dietary intake (5,79). Furthermore, rodents are relatively resistant to fructose ROLE OF URIC ACID IN INSULIN RESISTANCE AND FATTY in part because they generate less uric acid in response to LIVER IN HUMANS fructose due to the presence of the uricase gene in their As mentioned, fructose increases intracellular and circu- liver (38). Uricase degrades uric acid to allantoin, and as lating uric acid levels due to increased nucleotide turnover a consequence, rats degrade uric acid rapidly after it is

3310 DIABETES, VOL. 62, OCTOBER 2013 diabetes.diabetesjournals.org R.J. JOHNSON AND ASSOCIATES

TABLE 1 Serum uric acid predicts the development of diabetes Location Population End point F/U Independent Year First Author Israel 10,000 men diabetes 5 years Yes 1975 Medalie U.S. 5,209 adults diabetes 26 years Men 1985 Brand Naura 266 adults diabetes 6 years Women 1985 Balkau Sweden 766 men type 2 diabetes 13.5 years Yes 1988 Ohlson Britain 7,735 men NIDDM 12.8 years Yes 1995 Perry Kinmen (China) 654 high risk diabetes 3 years Yes 1998 Chou Mauritius 2,605 adults IGT or diabetes 5 years Yes 2000 Boyko Japan 6,356 men type 2 diabetes 9 years No 2001 Taniguchi Germany 6,166 adults type 2 diabetes 3–14 years Women 2002 Meisinger U.S. 9,020 adults ↑insulin 11 years Yes 2003 Carnethon Japan 2,310 men IFG or diabetes 6 years Yes 2003 Nakanishi Kinmen (China) 641 adults IFG or diabetes 7 years Women 2004 Lin U.S. 60 adults with MI ↑insulin 6 months Yes 2005 Nakagawa Finland 522 high risk type 2 diabetes 4.1 years No 2006 Niskanen Netherlands 4,536 adults type 2 diabetes 10 years Yes 2008 Dehghan Mauritius 4,259 adults diabetes 5 years Men 2008 Nan China 2,609 adults type 2 diabetes 9 years Yes 2008 Chien U.S. 9,689 adults met syn 5.7 years Yes 2008 Sui U.S. 566 high risk type 2 diabetes 13 years Yes 2009 Kramer U.S. 9,175 adults type 2 diabetes 26–28 years Yes 2010 Bhole Korea 4,779 men met syn 3 years Yes 2011 Ryu Japan 12,643 adults IFG and diabetes 5 years Women 2011 Yamada China 924 adults type 2 diabetes 3.5 years Yes 2011 Wang Italy 758 adults/BP type 2 diabetes 3 years Yes 2011 Viazzi ↑Insulin, hyperinsulinemia; BP, hypertension; F/U, follow-up; IFG, impaired fasting glucose; IGT, impaired glucose tolerance test; met syn, metabolic syndrome; MI, myocardial infarction; NIDDM, noninsulin-dependent diabetes mellitus. formed in their liver. When uricase is inhibited, rats show Expression of GLUT5 and the enzyme KHK, however, are a greater metabolic response to fructose with worse fatty enhanced with repeated exposure to fructose. It is in- liver and higher blood pressure (79). Indeed, there is evi- teresting that studies in children have found an inverse dence that the loss of uricase may have provided a survival relationship between fructose malabsorption and obesity advantage to ancestral apes living in Europe in the mid- (86). Consistent with this data, the metabolic response to Miocene and therefore may have acted like a thrifty gene fructose in children with NAFLD is greater compared with (80). The subsequent rise in sugar intake over the last lean control subjects (87). The importance of fructose centuries may have acted in concert with the loss of uri- absorption has recently been highlighted in African case to predispose us to obesity and diabetes (80). Americans because they commonly malabsorb fructose Clinical studies: the importance of the control group. and also have a lower frequency of NAFLD (88). The ob- Recently, a number of investigators have presented meta- servation that NAFLD subjects may absorb fructose more analyses that suggest fructose does not have a causal re- efficiently is further supported by our observation of lationship with obesity or metabolic syndrome (81–83). higher KHK expression in liver biopsies of NAFLD com- Before we analyze these studies, it is important to un- pared with other liver disease (36) and could be the reason derstand the complexity related to their interpretation. why ATP depletion in response to fructose is greater in First, many clinical studies use fructose alone—and often NAFLD subjects with a higher prior exposure to dietary at relatively high doses—in order to evaluate the effects of fructose (37). Our observation that hyperuricemia may fructose per se. This allows one to directly address the regulate KHK (35) also provides an explanation for why effects of fructose, and the use of high doses is a common studies in which fructose is given to young athletic lean experimental approach to allow one to identify metabolic individuals are often negative and why they may not carry effects that could otherwise take much longer periods to over to older and heavier individuals. show. Indeed, the fact that metabolic syndrome could be Another key issue is whether studies evaluating fruc- induced de novo in 25% of healthy men with high doses of tose should include fructose from natural fruits. One can fructose in just 2 weeks is a statement of how strong this argue that fructose is fructose regardless of source, but approach can be (63). Although studies involving HFCS or natural fruits also contain numerous substances that sucrose might be clinically more relevant, these types of block fructose effects, including potassium, vitamin C, studies will have trouble distinguishing whether the met- and antioxidants such as resveratrol, quercetin, and abolic effects observed are from the fructose or the high other flavonols. We found, for example, that whereas glycemic content of these added sugars. fructose from added sugars is associated with hyper- Nevertheless, the administration of fructose alone can tension, fructose from naturalfruitsisnot(89).Wefur- be very difficult to interpret because the absorption of ther showed that caloric restriction involving a reduction fructose when given alone is quite variable. As many as in fructose intake from added sugars could markedly two-thirds of children and one-third of adults malabsorb improve metabolic syndrome in obese Mexican adults, fructose (84,85). This is likely because of variable ex- and that this occurred even if natural fruits were pression of the fructose transporter GLUT5 in the gut. administered (90). diabetes.diabetesjournals.org DIABETES, VOL. 62, OCTOBER 2013 3311 FRUCTOSE, URIC ACID, AND DIABETES

FIG. 3. Uric acid: potential mechanisms for insulin resistance and diabetes. Uric acid may contribute to insulin resistance in the liver by inducing mitochondrial oxidative stress and steatosis (28). Uric acid also blocks the ability of insulin to stimulate vasodilation of blood vessels, which is important for the delivery of glucose to the skeletal muscle (4,32). Uric acid also induces local inflammation in the adipose tissue with a reduction in the production of adiponectin (44). Finally, uric acid may also have direct effects on the islet cells leading to local oxidative stress and islet dysfunction (5). Mt, mitochondria; PO4, phosphate.

Another important issue is whether glucose itself is the Blood pressure. It is a scientific fact that the adminis- right control for fructose. Outwardly it would seem true, tration of clinically relevant doses (60 g) of fructose but we recently discovered that glucose may act to induce acutely raises blood pressure in humans (92), and similar obesity and insulin resistance by being converted to fruc- increases in blood pressure have been observed following tose in the liver (91). Specifically, high concentrations of ingestion of 24 ounces of HFCS or sucrose-containing glucose, such as occurs in soft drinks, can induce the ac- beverages (68). It has also been reported that high doses of tivation of the polyol pathway in the liver, resulting in the fructose raises 24-h ambulatory blood pressure in humans and generation of fructose. In turn, the fructose is then me- can be blocked by lowering uric acid with allopurinol (63). tabolized by KHK, resulting in fructose-dependent effects. However, the recent meta-analysis by Ha et al. (82) addressed Indeed, glucose-induced weight gain, fat accumulation, whether short-term isocaloric fructose diets can increase fatty liver, and insulin resistance are all dependent on blood pressure after an overnight fast. Since the acute effects KHK. While some visceral fat and weight gain occur in of fructose to raise blood pressure occur during the ingestion glucose-fed mice lacking KHK, the development of fatty of fructose (and are likely mediated by uric acid), it is not liver and hyperinsulinemia are almost entirely dependent surprising that the authors did not show an effect on blood on glucose-induced fructose metabolism (91). Hence, al- pressure; indeed, a similarly designed study would conclude though fructose itself will have more metabolic effects that glucose-rich diets do not increase insulin levels. than glucose, the glucose itself may also be inducing An important question is whether chronic fructose in- metabolic changes via fructose. gestion may be responsible for persistent elevations in Meta-analyses that argue fructose is not a risk factor blood pressure. It is known that the greatest risk for per- for metabolic syndrome sistent hypertension is borderline hypertension in which Weight gain. A meta-analysis recently reported that fruc- intermittent blood pressure elevations occur. There is also tose intake does not cause weight gain compared with evidence that fructose causes microvascular disease in the other sugars in short-term studies if both groups are given kidney, which is known to predispose to persistent salt- the same number of total calories (isocaloric diets) (81). sensitive hypertension. Indeed, persistent hypertension However, no food will cause weight gain under these can be induced with fructose and high-salt diet in rats. conditions, as weight gain is driven primarily by increased Furthermore, chronic fructose ingestion over time is as- energy intake as opposed to a reduction in metabolic rate, sociated with elevations in fasting uric acid levels (58,93), at least in the short-term. Indeed, the mechanism by which in part because fructose also stimulates uric acid synthe- fructose increases weight is likely via its ability to stimu- sis. Epidemiological studies have also linked fructose in- late hunger and block satiety responses (7,9), so if food take with hypertension and elevated serum uric acid levels intake is controlled this would not be observed. (94). Reduction in sugar intake is also strongly associated

3312 DIABETES, VOL. 62, OCTOBER 2013 diabetes.diabetesjournals.org R.J. JOHNSON AND ASSOCIATES with a reduction in blood pressure (95). Indeed, the DASH ACKNOWLEDGMENTS (Dietary Approaches to Stop Hypertension) diet is in es- Support for this study was provided by ADA Basic Science sence a diet low in fructose from added sugars (while con- Award 7-12-BS-16 to Y.Y.S. taining natural fruits, see above). R.J.J. has patent applications related to lowering uric Uric acid. Wang et al. (83) also reported that short-term acid as a means to prevent or treat insulin resistance and isocaloric trials do not show an effect of fructose on features of metabolic syndrome; holds stock in XORT fasting uric acid levels. Again, the design of the study Pharma Corp.; and is on the Scientific Board of Amway. would not be expected to show a rise in uric acid because T.N. has patent applications related to lowering uric acid the initial rise in uric acid is transient and occurs within as a means to prevent or treat insulin resistance and minutes of the ingestion of fructose. However, as men- features of metabolic syndrome and holds stock in XORT tioned, there is some evidence that over time continued Pharma Corp. No other potential conflicts of interest ingestion of fructose will result in chronic elevations of relevant to this article were reported. uric acid. A more detailed discussion is provided elsewhere. R.J.J. wrote the first draft of the manuscript. T.N., Another issue with all three metanalyses is that they L.G.S.-L., M.S., S.S., M.L., T.I., Y.Y.S., and M.A.L. im- included control groups that ingested sucrose, which can proved the manuscript with their comments and sug- be questioned because sucrose is a disaccharide that gestions. In addition, alloftheauthorshavemade contains fructose (81–83). scientificcontributionsthatprovidedthegroundwork Other issues related to uric acid. Other aspects of the for this Perspective. uric acid studies have also been questioned. One paradox is that the acute elevation of serum uric acid by infusion REFERENCES often results in an improvement in endothelial function 1. Johnson RJ, Perez-Pozo SE, Sautin YY, et al. Hypothesis: could excessive (96). However, while uric acid is an antioxidant in the fructose intake and uric acid cause type 2 diabetes? Endocr Rev 2009;30: extracellular environment, it has prooxidative effects in- 96–116 side the cell (28,29). Several investigators have also sug- 2. Sánchez-Lozada LG, Mu W, Roncal C, et al. Comparison of free fructose gested that it is not uric acid that is driving metabolic and glucose to sucrose in the ability to cause fatty liver. Eur J Nutr 2010; syndrome, but rather xanthine oxidase, since this enzyme 49:1–9 generates oxidants in addition to uric acid, and it may be 3. Bocarsly ME, Powell ES, Avena NM, Hoebel BG. High-fructose corn syrup causes characteristics of obesity in rats: increased body weight, the former that is responsible for the metabolic syndrome. body fat and triglyceride levels. Pharmacol Biochem Behav 2010;97:101– For example, high-dose allopurinol improves endothelial 106 dysfunction in subjects with heart failure whereas the 4. Nakagawa T, Hu H, Zharikov S, et al. A causal role for uric acid in lowering of uric acid with probenecid was ineffective (97). fructose-induced metabolic syndrome. Am J Physiol Renal Physiol 2006; However, this could simply relate to the relative superi- 290:F625–F631 5. Roncal-Jimenez CA, Lanaspa MA, Rivard CJ, et al. Sucrose induces fatty ority of allopurinol to lower intracellular uric acid as it fl – liver and pancreatic in ammation in male breeder rats independent of blocks synthesis. Although xanthine oxidase induced excess energy intake. Metabolism 2011;60:1259–1270 oxidants could be important, the observation that raising 6. Teff KL, Elliott SS, Tschöp M, et al. Dietary fructose reduces circulating intracellular uric acid, even in the presence of allopurinol, insulin and leptin, attenuates postprandial suppression of ghrelin, and can increase hepatic fat suggests that it is the uric acid that increases triglycerides in women. J Clin Endocrinol Metab 2004;89:2963– is responsible (35). 2972 Genetic studies. A final argument relates to the genetics 7. Shapiro A, Mu W, Roncal C, Cheng KY, Johnson RJ, Scarpace PJ. Fructose-induced leptin resistance exacerbates weight gain in response of uric acid and fructose metabolism. While some genetic to subsequent high-fat feeding. Am J Physiol Regul Integr Comp Physiol polymorphisms in various enzymes involved in fructose 2008;295:R1370–R1375 metabolism and urate transport have been linked with 8. Bernal SY, Dostova I, Kest A, et al. Role of dopamine D1 and D2 receptors metabolic syndrome and hypertension (98–100), several in the nucleus accumbens shell on the acquisition and expression of genome-wide association studies (GWAS) could not show fructose-conditioned flavor-flavor preferences in rats. Behav Brain Res such associations (101,102). However, the primary poly- 2008;190:59–66 9. Lane MD, Cha SH. Effect of glucose and fructose on food intake via morphism driving serum uric acid in the GWAS studies malonyl-CoA signaling in the brain. Biochem Biophys Res Commun 2009; is SLC2A9;thispolymorphismmediatesthetransportof 382:1–5 uric acid out of tubular cells,andsoitmaynotpredict 10. Friedman MI, Harris RB, Ji H, Ramirez I, Tordoff MG. Fatty acid oxidation the development of diabetes because it is likely to dis- affects food intake by altering hepatic energy status. Am J Physiol 1999; sociate the serum from intracellular uric acid levels, the 276:R1046–R1053 latter of which may be more important in driving insulin 11. Bawden SJ, Stephenson MC, Marciani L, Aithal GP, Macdonald IA, Gowland P, Morris PA. Investigating alterations in hepatic atp levels resistance. following fructose and fructose+glucose ingestion: a simple non-invasive technique to assess liver function using 31P MRS. Proc Intl Soc Magn Reson Med Sci Meet Exhib 2012;20:1369 CONCLUSIONS 12. Cox CL, Stanhope KL, Schwarz JM, et al. Consumption of fructose- Searching for the cause of type 2 diabetes has been sweetened beverages for 10 weeks reduces net fat oxidation and energy a prime area of research since Etienne Lancereaux de- expenditure in overweight/obese men and women. Eur J Clin Nutr 2012; 66:201–208 scribed fat diabetes (diabetes gras) in 1880. Though more 13. Basu S, Yoffe P, Hills N, Lustig RH. The relationship of sugar to population- studies are needed, the evidence that fructose-induced hy- level diabetes prevalence: an econometric analysis of repeated cross- peruricemia may have a contributory role is gaining ground. sectional data. PLoS ONE 2013;8:e57873 While it still remains a hypothesis (103), increasing evi- 14. Mäenpää PH, Raivio KO, Kekomäki MP. Liver adenine nucleotides: dence suggests uric acid may have a fundamental role in the fructose-induced depletion and its effect on protein synthesis. Science manifestations of metabolic syndrome (Fig. 3). Given that 1968;161:1253–1254 15. van den Berghe G, Bronfman M, Vanneste R, Hers HG. The mechanism hyperuricemia is a remediable risk factor, we recommend of adenosine triphosphate depletion in the liver after a load of fructose. both basic and clinical studies to address this important Akineticstudyofliveradenylatedeaminase.BiochemJ1977;162:601– possibility. 609 diabetes.diabetesjournals.org DIABETES, VOL. 62, OCTOBER 2013 3313 FRUCTOSE, URIC ACID, AND DIABETES

16. Bode JC, Zelder O, Rumpelt HJ, Wittkamp U. Depletion of liver adenosine 41. Ryu S, Chang Y, Kim SG, Cho J, Guallar E. Serum uric acid levels predict phosphates and metabolic effects of intravenous infusion of fructose or incident nonalcoholic fatty liver disease in healthy Korean men. Metab- sorbitol in man and in the rat. Eur J Clin Invest 1973;3:436–441 olism 2011;60:860–866 17. Emmerson BT. Effect of oral fructose on urate production. Ann Rheum 42. Hoehn KL, Salmon AB, Hohnen-Behrens C, et al. Insulin resistance is Dis 1974;33:276–280 a cellular antioxidant defense mechanism. Proc Natl Acad Sci USA 2009; 18. Ishimoto T, Lanaspa MA, Le MT, et al. Opposing effects of fructokinase C 106:17787–17792 and A isoforms on fructose-induced metabolic syndrome in mice. Proc 43. Samuel VT, Petersen KF, Shulman GI. Lipid-induced insulin resistance: Natl Acad Sci USA 2012;109:4320–4325 unravelling the mechanism. Lancet 2010;375:2267–2277 19. Lanaspa MA, Sanchez-Lozada LG, Choi YJ, et al. Uric acid induces hepatic 44. Baldwin W, McRae S, Marek G, et al. Hyperuricemia as a mediator of the steatosis by generation of mitochondrial oxidative stress: potential role in proinflammatory endocrine imbalance in the adipose tissue in a murine fructose-dependent and -independent fatty liver. J Biol Chem 2012;287: model of the metabolic syndrome. Diabetes 2011;60:1258–1269 40732–40744 45. Lanaspa M, Sautin Y, Ejaz A, et al. Uric acid and metabolic syndrome: 20. Steinmann B, Gitzelmann R, Van den Berghe G. Disorders of Fructose what is the relationship? Curr Rheum Rev 2011;7:162–169 Metabolism. In The Metabolic and Molecular Basis of Inherited Disease. 46. Duplain H, Burcelin R, Sartori C, et al. Insulin resistance, hyperlipidemia, Scriver C, Beaudet A, Sly W, Valle D, Eds. New York, McGraw-Hill, 2001, and hypertension in mice lacking endothelial nitric oxide synthase. Cir- p. 1489–1520 culation 2001;104:342–345 21. Odièvre M, Gentil C, Gautier M, Alagille D. Hereditary fructose in- 47. Schwartz IF, Grupper A, Chernichovski T, et al. Hyperuricemia attenuates tolerance in childhood. Diagnosis, management, and course in 55 pa- aortic nitric oxide generation, through inhibition of arginine transport, in tients. Am J Dis Child 1978;132:605–608 rats. J Vasc Res 2011;48:252–260 22. Bandsma RH, Smit GP, Kuipers F. Disturbed lipid metabolism in glycogen 48. Zharikov S, Krotova K, Hu H, et al. Uric acid decreases NO production storage disease type 1. Eur J Pediatr 2002;161(Suppl. 1):S65–S69 and increases arginase activity in cultured pulmonary artery endothelial 23. Greene HL, Wilson FA, Hefferan P, et al. ATP depletion, a possible role in cells. Am J Physiol Cell Physiol 2008;295:C1183–C1190 the pathogenesis of hyperuricemia in glycogen storage disease type I. 49. Kang DH, Park SK, Lee IK, Johnson RJ. Uric acid-induced C-reactive J Clin Invest 1978;62:321–328 protein expression: implication on cell proliferation and nitric oxide 24. Mundy HR, Lee PJ. Glycogenosis type I and diabetes mellitus: a common production of human vascular cells. J Am Soc Nephrol 2005;16:3553–3562 mechanism for renal dysfunction? Med Hypotheses 2002;59:110–114 50. Khosla UM, Zharikov S, Finch JL, et al. Hyperuricemia induces endo- 25. Spiegel R, Rakover-Tenenbaum Y, Mandel H, Lumelski D, Admoni O, thelial dysfunction. Kidney Int 2005;67:1739–1742 Horovitz Y. Secondary diabetes mellitus: late complication of glycogen 51. Sánchez-Lozada LG, Soto V, Tapia E, et al. Role of oxidative stress in the storage disease type 1b. J Pediatr Endocrinol Metab 2005;18:617–619 renal abnormalities induced by experimental hyperuricemia. Am 26. Ouyang J, Parakhia RA, Ochs RS. Metformin activates AMP kinase J Physiol Renal Physiol 2008;295:F1134–F1141 through inhibition of AMP deaminase. J Biol Chem 2011;286:1–11 52. Sánchez-Lozada LG, Tapia E, López-Molina R, et al. Effects of acute 27. Lanaspa MA, Cicerchi C, Garcia G, et al. Counteracting roles of AMP and chronic L-arginine treatment in experimental hyperuricemia. Am deaminase and AMP kinase in the development of fatty liver. PLoS ONE J Physiol Renal Physiol 2007;292:F1238–F1244 2012;7:e48801 53. Kanbay M, Segal M, Afsar B, Kang DH, Rodriguez-Iturbe B, Johnson RJ. 28. Lanaspa MA, Sanchez-Lozada LG, Choi YJ, et al. Uric acid induces hepatic The role of uric acid in the pathogenesis of human cardiovascular dis- steatosis by generation of mitochondrial oxidative stress: potential role in ease. Heart 2013;99:759–766 fructose-dependent and -independent fatty liver. J Biol Chem 2012;287: 54. Cummings BP, Stanhope KL, Graham JL, et al. Dietary fructose accel- 40732–40744 erates the development of diabetes in UCD-T2DM rats: amelioration by 29. Sautin YY, Nakagawa T, Zharikov S, Johnson RJ. Adverse effects of the the antioxidant, alpha-lipoic acid. Am J Physiol Regul Integr Comp classic antioxidant uric acid in adipocytes: NADPH oxidase-mediated ox- Physiol 2010;298:R1343–R1350 idative/nitrosative stress. Am J Physiol Cell Physiol 2007;293:C584–C596 55. Malik VS, Popkin BM, Bray GA, Després JP, Willett WC, Hu FB. Sugar- 30. Yu MA, Sánchez-Lozada LG, Johnson RJ, Kang DH. Oxidative stress with sweetened beverages and risk of metabolic syndrome and type 2 di- an activation of the renin-angiotensin system in human vascular endo- abetes: a meta-analysis. Diabetes Care 2010;33:2477–2483 thelial cells as a novel mechanism of uric acid-induced endothelial dys- 56. Qi Q, Chu AY, Kang JH, et al. Sugar-sweetened beverages and genetic risk function. J Hypertens 2010;28:1234–1242 of obesity. N Engl J Med 2012;367:1387–1396 31. Corry DB, Eslami P, Yamamoto K, Nyby MD, Makino H, Tuck ML. Uric 57. Hallfrisch J, Reiser S, Prather ES. Blood lipid distribution of hyperinsulinemic acid stimulates vascular smooth muscle cell proliferation and oxidative stress men consuming three levels of fructose. Am J Clin Nutr 1983;37:740–748 via the vascular renin-angiotensin system. J Hypertens 2008;26:269–275 58. Reiser S, Powell AS, Scholfield DJ, Panda P, Ellwood KC, Canary JJ. 32. Sanchez-Lozada LG, Lanaspa-Garcia MA, Cristobal M, et al. Uric acid- Blood lipids, lipoproteins, apoproteins, and uric acid in men fed diets induced endothelial dysfunction is associated with mitochondrial alter- containing fructose or high-amylose cornstarch. Am J Clin Nutr 1989;49: ations and decreased intracellular ATP concentrations. Nephron Exp 832–839 Nephrol 2012;121:e71–e78 59. Stanhope KL, Schwarz JM, Keim NL, et al. Consuming fructose-sweetened, 33. Imaram W, Gersch C, Kim KM, Johnson RJ, Henderson GN, Angerhofer A. not glucose-sweetened, beverages increases visceral adiposity and lipids Radicals in the reaction between peroxynitrite and uric acid identified by and decreases insulin sensitivity in overweight/obese humans. J Clin Invest electron spin resonance spectroscopy and liquid chromatography mass 2009;119:1322–1334 spectrometry. Free Radic Biol Med 2010;49:275–281 60. Cox CL, Stanhope KL, Schwarz JM, et al. Consumption of fructose- but 34. Gersch C, Palii SP, Kim KM, Angerhofer A, Johnson RJ, Henderson GN. not glucose-sweetened beverages for 10 weeks increases circulating Inactivation of nitric oxide by uric acid. Nucleosides Nucleotides Nucleic concentrations of uric acid, retinol binding protein-4, and gamma-glutamyl Acids 2008;27:967–978 transferase activity in overweight/obese humans. Nutr Metab (Lond) 2012; 35. Lanaspa MA, Sanchez-Lozada LG, Cicerchi C, et al. Uric acid stimulates 9:68 fructokinase and accelerates fructose metabolism in the development of 61. Maersk M, Belza A, Stødkilde-Jørgensen H, et al. Sucrose-sweetened fatty liver. PLoS ONE 2012;7:e47948 beverages increase fat storage in the liver, muscle, and visceral fat depot: 36. Ouyang X, Cirillo P, Sautin Y, et al. Fructose consumption as a risk factor a 6-mo randomized intervention study. Am J Clin Nutr 2012;95:283–289 for non-alcoholic fatty liver disease. J Hepatol 2008;48:993–999 62. Lê KA, Ith M, Kreis R, et al. Fructose overconsumption causes dyslipi- 37. Abdelmalek MF, Lazo M, Horska A, et al.; Fatty Liver Subgroup of Look demia and ectopic lipid deposition in healthy subjects with and without AHEAD Research Group. Higher dietary fructose is associated with im- a family history of type 2 diabetes. Am J Clin Nutr 2009;89:1760–1765 paired hepatic adenosine triphosphate homeostasis in obese individuals 63. Perez-Pozo SE, Schold J, Nakagawa T, Sánchez-Lozada LG, Johnson RJ, with type 2 diabetes. Hepatology 2012;56:952–960 Lillo JL. Excessive fructose intake induces the features of metabolic 38. Stavric B, Johnson WJ, Clayman S, Gadd RE, Chartrand A. Effect of syndrome in healthy adult men: role of uric acid in the hypertensive re- fructose administration on serum urate levels in the uricase inhibited rat. sponse. Int J Obes (Lond) 2010;34:454–461 Experientia 1976;32:373–374 64. de Ruyter JC, Olthof MR, Seidell JC, Katan MB. A trial of sugar-free or 39. Xu CF, Yu CH, Xu L, Sa XY, Li YM. Hypouricemic therapy: a novel po- sugar-sweetened beverages and body weight in children. N Engl J Med tential therapeutic option for nonalcoholic fatty liver disease. Hepatology 2012;367:1397–1406 2010;52:1865–1866 65. Gardner CD, Kiazand A, Alhassan S, et al. Comparison of the Atkins, 40. Kono H, Rusyn I, Bradford BU, Connor HD, Mason RP, Thurman RG. Zone, Ornish, and LEARN diets for change in weight and related risk Allopurinol prevents early alcohol-induced liver injury in rats. J Phar- factors among overweight premenopausal women: the A TO Z Weight macol Exp Ther 2000;293:296–303 Loss Study: a randomized trial. JAMA 2007;297:969–977

3314 DIABETES, VOL. 62, OCTOBER 2013 diabetes.diabetesjournals.org R.J. JOHNSON AND ASSOCIATES

66. James J, Thomas P, Cavan D, Kerr D. Preventing childhood obesity by 86. Disse S, Buelow A, Boedeker RH, et al. Reduced prevalence of obesity in reducing consumption of carbonated drinks: cluster randomised con- children with primary fructose malabsorption: a multicenter, retrospec- trolled trial. BMJ 2004;328:1237 tive cohort study. Pediatr Obes 2013;8:255–258 67. Shi L, van Meijgaard J. Substantial decline in sugar-sweetened beverage 87. Jin R, Le NA, Liu S, et al. Children with NAFLD are more sensitive to the consumption among California’s children and adolescents. Int J Gen Med adverse metabolic effects of fructose beverages than children without 2010;3:221–224 NAFLD. J Clin Endocrinol Metab 2012;97:E1088–E1098 68. Le MT, Frye RF, Rivard CJ, et al. Effects of high-fructose corn syrup and 88. Walker RW, Lê KA, Davis J, et al. High rates of fructose malabsorption are sucrose on the pharmacokinetics of fructose and acute metabolic and associated with reduced liver fat in obese African Americans. J Am Coll hemodynamic responses in healthy subjects. Metabolism 2012;61:641–651 Nutr 2012;31:369–374 69. Choi JW, Ford ES, Gao X, Choi HK. Sugar-sweetened soft drinks, diet soft 89. Madero M, Arriaga JC, Jalal D, et al. The effect of two energy-restricted drinks, and serum uric acid level: the Third National Health and Nutrition diets, a low-fructose diet versus a moderate natural fructose diet, on Examination Survey. Arthritis Rheum 2008;59:109–116 weight loss and metabolic syndrome parameters: a randomized con- 70. Masuo K, Kawaguchi H, Mikami H, Ogihara T, Tuck ML. Serum uric acid trolled trial. Metabolism 2011;60:1551–1559 and plasma norepinephrine concentrations predict subsequent weight 90. Jalal DI, Smits G, Johnson RJ, Chonchol M. Increased fructose associates gain and blood pressure elevation. Hypertension 2003;42:474–480 with elevated blood pressure. J Am Soc Nephrol 2010;21:1543–1549 71. Feig DI, Madero M, Jalal DI, Sanchez-Lozada LG, Johnson RJ. Uric acid 91. Lanaspa MA, Ishimoto T, Li N, et al. Glucose-induced obesity, fatty liver and the origins of hypertension. J Pediatr 2013;162:896–902 and insulin resistance is mediated by endogenous fructose. Nature 72. Ruggiero C, Cherubini A, Miller E 3rd, et al. Usefulness of uric acid to Commun. In press predict changes in C-reactive protein and interleukin-6 in 3-year period in 92. Brown CM, Dulloo AG, Yepuri G, Montani JP. Fructose ingestion acutely Italians aged 21 to 98 years. Am J Cardiol 2007;100:115–121 elevates blood pressure in healthy young humans. Am J Physiol Regul 73. Choi HK, Ford ES. Prevalence of the metabolic syndrome in individuals Integr Comp Physiol 2008;294:R730–R737 with hyperuricemia. Am J Med 2007;120:442–447 93. Hallfrisch J, Ellwood K, Michaelis OE 4th, Reiser S, Prather ES. Plasma 74. Facchini F, Chen YD, Hollenbeck CB, Reaven GM. Relationship between fructose, uric acid, and inorganic phosphorus responses of hyper- resistance to insulin-mediated glucose uptake, urinary uric acid clear- insulinemic men fed fructose. J Am Coll Nutr 1986;5:61–68 ance, and plasma uric acid concentration. JAMA 1991;266:3008–3011 94. Nguyen S, Choi HK, Lustig RH, Hsu CY. Sugar-sweetened beverages, 75. Reungjui S, Roncal CA, Mu W, et al. Thiazide diuretics exacerbate fructose- serum uric acid, and blood pressure in adolescents. J Pediatr 2009;154: induced metabolic syndrome. J Am Soc Nephrol 2007;18:2724–2731 807–813 76. Ogino K, Kato M, Furuse Y, et al. Uric acid-lowering treatment with 95. Chen L, Caballero B, Mitchell DC, et al. Reducing consumption of sugar- benzbromarone in patients with heart failure: a double-blind placebo- sweetened beverages is associated with reduced blood pressure: a pro- controlled crossover preliminary study. Circ Heart Fail 2010;3:73–81 spective study among United States adults. Circulation 2010;121: 77. Dogan A, Yarlioglues M, Kaya MG, et al. Effect of long-term and high-dose 2398–2406 allopurinol therapy on endothelial function in normotensive diabetic pa- 96. Waring WS, McKnight JA, Webb DJ, Maxwell SR. Uric acid restores en- tients. Blood Press 2011;20:182–187 dothelial function in patients with type 1 diabetes and regular smokers. 78. Truswell AS, Seach JM, Thorburn AW. Incomplete absorption of pure Diabetes 2006;55:3127–3132 fructose in healthy subjects and the facilitating effect of glucose. Am J 97. George J, Carr E, Davies J, Belch JJ, Struthers A. High-dose allopurinol Clin Nutr 1988;48:1424–1430 improves endothelial function by profoundly reducing vascular oxidative 79. Tapia E, Cristóbal M, García-Arroyo FE, et al. Synergistic effect of uricase stress and not by lowering uric acid. Circulation 2006;114:2508–2516 blockade plus physiological amounts of fructose-glucose on glomerular 98. Shafiu M, Johnson RJ, Turner ST, et al. Urate transporter gene SLC22A12 hypertension and oxidative stress in rats. Am J Physiol Renal Physiol polymorphisms associated with obesity and metabolic syndrome in 2013;304:F727–F736 Caucasians with hypertension. Kidney Blood Press Res 2012;35:477–482 80. Johnson RJ, Andrews P. Fructose, uricase, and the back-to-Africa Hy- 99. Parsa A, Brown E, Weir MR, et al. Genotype-based changes in serum uric pothesis. Evol Anthropol 2010;19:250–257 acid affect blood pressure. Kidney Int 2012;81:502–507 81. Sievenpiper JL, de Souza RJ, Mirrahimi A, et al. Effect of fructose on body 100. Le MT, Lobmeyer MT, Campbell M, et al. Impact of genetic poly- weight in controlled feeding trials: a systematic review and meta-analysis. morphisms of SLC2A2, SLC2A5, and KHK on metabolic phenotypes in Ann Intern Med 2012;156:291–304 hypertensive individuals. PLoS ONE 2013;8:e52062 82. Ha V, Sievenpiper JL, de Souza RJ, et al. Effect of fructose on blood 101. Pfister R, Barnes D, Luben R, et al. No evidence for a causal link between pressure: a systematic review and meta-analysis of controlled feeding uric acid and type 2 diabetes: a Mendelian randomisation approach. Di- trials. Hypertension 2012;59:787–795 abetologia 2011;54:2561–2569 83. Wang DD, Sievenpiper JL, de Souza RJ, et al. The effects of fructose intake on 102. Yang Q, Köttgen A, Dehghan A, et al. Multiple genetic loci influence serum uric acid vary among controlled dietary trials. J Nutr 2012;142:916–923 serum urate levels and their relationship with gout and cardiovascular 84. Ravich WJ, Bayless TM, Thomas M. Fructose: incomplete intestinal ab- disease risk factors. Circ Cardiovasc Genet 2010;3:523–530 sorption in humans. Gastroenterology 1983;84:26–29 103. Nakagawa T, Tuttle KR, Short RA, Johnson RJ. Hypothesis: fructose- 85. Kneepkens CM, Vonk RJ, Fernandes J. Incomplete intestinal absorption induced hyperuricemia as a causal mechanism for the epidemic of the of fructose. Arch Dis Child 1984;59:735–738 metabolic syndrome. Nat Clin Pract Nephrol 2005;1:80–86

diabetes.diabetesjournals.org DIABETES, VOL. 62, OCTOBER 2013 3315